1) There's too much noise on the new page.
2) It matters which at what time of day you submit an article.
3) There's too much chance about whether it'll been seen or not.
4) Someone else will post it with a slightly different URL (see above), and that's the one which will be noticed.
I am, of course, very grateful for everyone else who posts stuff. There's usually a good article or two on the front page.
I suppose the question we should ask is if it matters if decent articles are missed? There's lots of text out there on the internet.
[edit: Removed three trivial ways to defeat the dup detector.]
Personally, I think a separate section is needed just for HN'ers to showcase/demo/launch their startups or projects - "Show HN" threads, for the most part, have even less chance of getting eyeballs especially if they're self-posts; plus, there's the whole luck/timing thing as well. Or instead of a separate section, perhaps they should be given more weight, considering...isn't that what HN is all about in the first place?
I second the suggestion for a separate section for the "Show HN" posts, as it would give each post a better chance of being seen by those willing to provide feedback.
Wouldn't it be kinda neat to have a webapp that locates all of the HN submissions with just 1 lonely point from the past day, and picks one at random to feature prominently?
Along similar lines, last week I whipped up a little tool called Momentum that lets you "pre-promote" links that you plan to submit to HN, so that people who think a link is interesting can plan to up-vote it when it's actually submitted:
Ironically, I submitted Momentum to HN ... and it never gained traction.
The question is rather: how do make sure the average quality of the articles on the front page is as high as possible?
Probably we're quite close to as good as it gets.
One option would be to have a "runner ups"-page ("bubblers"?) which lists the newly submitted articles that are _almost_ on the front page. So instead of two pages (front and new) we would have three (front, bubblers, new).
The algorithm would be simple: the 30 latest articles with more than 1 point that are not yet on the front page.
Hm. This actually sounds like quite a fun little weekend project. Anyone up for it? Is almosthn.com available? :)
Ironically, I've felt this way for a while and created a website to test this theory.
And still better, when a thread gets archived (when the comment thread becomes read-only; happens after some N days), add its current score delta (the number of points it has now, minus the number it had when you initially invested yours) back into the user's karma, times some coefficient (possibly just 1.)
And finally, allow anyone to do this (invest karma; reap eventual score-delta rewards), not just the initial poster.
Bam: average-sized link futures market.
remove tongue from cheek
Seriously, why? This would seem to me to be the perfect way to breed people who care more about karma than content... and they're the problem, IMHO.
[1] For some definition of 'all'.
For example, it could look like this:
My overlooked submissions list, from newest to oldest (last updated: xx.xx.xxxx):
1. Article Name (http://domain.com)
2.
3.
4.
5.
I think that you can give trusted members in the community more weight when voting. Determination shouldnt be done in an automated way.